A Semantic Interoperability Framework for Data-Centric Applications in Agriculture

The rapid growth of data-centric applications in agriculture has generated vast and heterogeneous datasets, yet their potential is constrained by the lack of semantic interoperability, which limits meaningful data exchange, integration, and reuse. This dissertation proposes a Semantic Interoperability Framework that integrates metadata schemas, ontologies, knowledge graphs, and artificial intelligence to resolve interoperability conflicts in naming conventions, domain representation, and metadata alignment, while adhering to the FAIR data principles. Developed through a Design Science Research methodology, the framework combines structured metadata annotation, ontological modeling for semantic alignment, and knowledge graph construction to enhance data linking and reasoning, with a large language model (LLM) supporting knowledge graph generation and creating SPARQL queries from natural language prompts. Its applicability is demonstrated through two case studies: (1) Agricultural Price Index Data in Brazil, which aligns datasets from CEPEA, IPEA, and CONAB using the Almes Core metadata schema and the APTO ontology for agricultural product types; and (2) Agrobiodiversity and Plant–Pollinator Interaction Data within the WorldFAIR project, which shows how FAIR-aligned schemas and ontology-driven integration standardize complex ecological datasets for scientific collaboration. Evaluation through ontology validation metrics, usability testing, and query performance demonstrates significant improvements in data interoperability, enabling more accurate retrieval, integration, and machine-driven reasoning. The findings also highlight persistent challenges such as metadata adoption, automation of ontology construction, and the need for stakeholder engagement in standardization efforts. Overall, this research offers a novel, scalable, and reusable approach to achieving semantic interoperability in agriculture, bridging fragmented datasets and advancing open data initiatives, digital agriculture policies, and AI-driven analytics.